Background
Human papillomaviruses (HPVs) cause one of the most common sexually transmitted infections in the world. A subset of "high-risk" HPV genotypes is unequivocally associated to cervical cancer, the second main cause of death for cancer in women worldwide [
1,
2]. Nowadays the impending commercialisation of the prophylactic anti-HPV vaccine is shifting research efforts towards tumour therapy. Many efforts have been made to develop effective treatments for the HPV-associated lesions [
3]. The delivery of antitumoral agents to the cervical cancer cells may represent a valid strategy for their treatment especially at an early lesion stage, and in addition or alternative to surgery of the advanced lesions.
The viral proteins E6 and E7, which play a crucial role in viral oncogenesis [
4‐
6] are recognised "tumor-specific antigens" and are therefore considered suitable targets for either immunotherapy or therapeutic vaccination against the HPV-associated tumors [
7].
Recombinant single-chain variable fragments (scFv) antibodies represent powerful tools for different immunotherapy purposes and are particularly suitable in intracellular immunisation to knock out specific protein functions [
8‐
11]. Many of them are either in clinical trial or successfully used in therapy.
The good potential of scFvs for biomedical use is severely limited by their intrinsic solubility and stability, which are important characteristics to achieve long-lasting effects both
in vitro and
in vivo [
12]. However, it is worth noting that the
in vitro and
in vivo scFv specificities are not always comparable to one another. ScFv solubility and stability are related to their primary structure and mostly depend on the intrinsic ability of correctly folding by forming intra-chain disulphide bonds in reducing environments, in both prokaryotic and eukaryotic cells [
13,
14]. The scFv thermal stability is a decisive property for their applications to targeted tumor therapy [
15]. In fact the scFvs must be stable at 37°C for many hours to be able to penetrate into tumors, an activity that can take 12 hours or more [
16].
We have previously reported the selection of three different scFvs (scFv 32, 43 and 51) against the E7 oncoprotein of the HPV16 (16E7) from the ETH-2 phage display library of human antibody fragments [
17]. We have described the antiproliferative effect of the most reactive scFv43, when it is expressed in the nuclear and secretory compartments of the HPV16-positive cervical carcinoma SiHa cell line [
18]. We have also demonstrated that this effect could be specifically ascribed to inhibition of the E7 activity, in promoting cell proliferation.
In this paper we have investigated the biophysical properties of the three anti-E7 scFvs when these are expressed in either prokaryotic or eukaryotic systems, in the attempt to focus parameters important for their activity both
in vitro and
in vivo [
19].
To establish if the differences observed in the scFv stability could be related to their secondary structure, we have compared the scFv amino acid sequences to the consensus sequences of both the IMGT database and the validated intrabody database (VIDA) of intracellular stable scFvs [
20,
21]. Two main non-synonymous and possibly destabilizing mutations were identified in the scFv43 sequence. A mutagenesis strategy was designed to partially revert these mutations back to the original sequence, obtaining 2 novel antibodies, scFv43 M1 and scFv43 M2.
The 5 anti-16E7 scFvs, 2 obtained by mutagenesis and 3 originally selected, were compared to one another in terms of thermal stability and solubility in both prokaryotic and eukaryotic systems.
Methods
ScFvs selection
The anti-HPV16E7 scFvs were selected from the ETH-2 phage display library of human antibody fragments after three rounds of panning in solution against the recombinant His-HPV16E7 protein as already described by Accardi et al. [
18].
Cloning of scFv32 and 51 sequences in scFvExpress vectors and plasmids
The scFv32 and 51 coding sequence were isolated from pDN332 by NcoI/NotI restriction and subcloned into the vectors scFvE-cyto and scFvE-nuclear, digested with the same enzymes. For cloning into scFvE-sekdel, the encoding sequences were PCR-amplified (95°C for 1 minute, 50°C for 1 minute, 74°C for 1 minute, 35 cycles) using the SalsekD-NotsekR2 couple of primers. The sequences of the primers used are the following (restriction sites are underlined):
SalsekD 5'CGGCGTCGACCCGAGGTGCAGCTGGTGG 3'
NotsekR2 5'CGGCGCGGCCGCTTTGATTTCCACCTTGGTCCC 3'
PCR products were double-digested with SalI/NotI, gel-purified using GFX PCR DNA and Gel Band Purification Kit (Amersham Biosciences) and ligated into scFvE-sekdel digested with the same enzymes. The ligation products were used to transform competent XL1-blue E. coli cells (Stratagene). The scFv fragments isolated with the described procedure and cloned into the scFvExpress vectors do not retain the FLAG-tag and the His-tag sequences present in the original plasmid, and acquire a c-myc-tag at their C-terminus, utilised for detection. The R4-cyto construct expressing an anti-β-gal scFv targeted to the cytoplasm, utilised as a control of stable and soluble scFv antibody, was kindly provided by A. Cattaneo.
Sequence alignments
Sequence alignments to NCBI and to VIDA were carried out using Immunoglobulin BLAST and CLUSTALL programs.
Protein expression and purification from periplasmic and total extracts
The His-E7 expression and purification from the JM109
E. coli strain and the scFv expression from the HB2151
E. coli strain have been already described by Accardi et al. [
18]. Briefly, 1 ml-aliquots of each bacterial culture were collected just before isopropylthio-β-galactoside (IPTG) induction (non induced samples), and at 2–4 hours post-induction, pelletted and resuspended in 100 μl of 2× SDS-loading buffer. For each sample, 20 μl-aliquots of both the pellet and the supernatant were analysed for the scFv expression by Western blotting followed by chemiluminescence as described below. For scFv expression in the M15
E. coli strain, an overnight (ON) culture was diluted 1: 40 in 2xTY 2%, glucose/amp (100 μg/ml)/Kan (25 μg/ml), grown until OD
600= 0.8, and centrifuged to harvest bacteria. The pellet was recovered in 1l of 2xTY/amp/2 mM IPTG and the culture grown for 4 hours at room temperature (RT). The bacteria recovered from this culture after centrifugations were treated differently to prepare periplasmic extract (PE) and total extract (TE). For PE preparation, the bacteria were resuspended in 50 ml TES (Tris-HCl pH = 7, 20% saccarose, 1 mM EDTA) in the presence of protease inhibitors (Roche). After centrifugation at 4000 rpm for 40' at 4°C, the supernatant (PE1) was kept at 4°C and the pellet resuspended in 50 ml of MgSO
4, shaken 10' at RT, and centrifuged at 10000 rpm for 20' at 4°C; the supernatant (PE2) was harvested after addition of protease inhibitors and combined to PE1, obtaining PE. TE extracts were prepared by resuspending the bacteria recovered from 1 litre-culture in 50 ml of lysis buffer (50 mM Tris-HCl pH = 7, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 0.5 M NaCl) containing Complete EDTA-free protease inhibitors (Roche), sonicated and incubated for 20' on ice. After centrifugation at 12,000 g for 10' at 4°C, supernatant (TE) was recovered and used for the scFv purification. The scFv purification was performed by affinity chromatography either on Ni-NTA beads (Qiagen) or on protein A-Sepharose CL-4B (Amersham Biosciences), according to the manufacturers' instructions. Purity of the proteins was evaluated by Coomassie Blue Staining after SDS-PAGE.
ScFv reactivity and affinity determination
The scFv reactivity and affinity for 16E7 were evaluated by Western blot analysis and ELISA as already described by Accardi et al. [
18]. Briefly, the competitive ELISA for affinity determination was performed by incubating in solution the purified scFvs at concentration of 0.7 μg/ml, with His-E7 (competitor) at different concentrations in the range 10
-10 and 10
-6 M, for 2 hours. The samples were then incubated ON in microtiter wells coated with 300 ng/well His-E7, at 4°C; the reaction was revealed by the anti-FLAG M2 mAb (Sigma St. Louis, MO) followed by GAM-HRP IgG (Amresco, Solon, OH) and using TMB substrate kit for peroxidase (Vector Laboratories, Inc. Burlingame, CA) colorimetric evaluation. The affinity constant was defined as the reciprocal of the competitor concentration required to reduce of the 50% the binding of the antibody not preincubated with E7, and is expressed in M
-1.
ScFv thermal stability
Purified scFvs were diluted in Dulbecco's PBS (DPBS) + Ca2+ + Mg2+ containing 0.2% human serum albumin (HSA), and incubated either at 37°C for predefined intervals of time or at 40°-50°-60°C for 10'. After incubation, the antibodies were assayed for their ability of binding the E7 protein by ELISA. Coating of a microtiter 96-wells plate (Nunc Polysorp) was performed ON at 4°C with purified recombinant His-E7 at 300 ng/well in carbonate/bicarbonate buffer (Pierce, Rockford, IL). The plate was saturated with 2% NFDM and then incubated with the purified scFvs at the desired dilutions ON at 4°C. Immunocomplexes were detected using the anti-FLAG M2 mAb (Sigma, St. Louis, MO) at the dilution 1:3000, followed by GAM-HRP IgG (Amresco, Solon, OH) as above described.
ScFv43 mutagenesis
Mutations in scFv43 were obtained by PCR amplification of the scFv43 sequence using QuickChange™ Site-Directed Mutagenesis Kit (Stratagene) and the following primers (the mutated nucleotides are underlined):
43M1 5' ctgtcagcagcgtcatggtaatccggc 3'
43M1 5' gccggattaccatgacgctgctgacag 5'
43M2 5' gcagctatgccacgagctgggtccgcc 3'
43M2 5' ggcggacccagctcgtggcatagctgc 3'
The PCR reaction was performed as follows: 95°C for 30 seconds, 45°C for 1 minute; 68°C for 9 minutes, 18 cycles. The PCR products were used to transform competent DH5α cells (Stratagene). Plasmids DNA from the obtained clones as extracted using Qiagen Spin Kit (Qiagen) and sequenced to confirm the incorporation of the desired mutations. ScFv43 M1 and scFv43 M2 plasmids were used to transform competent HB2151 cells (Stratagene) for scFv production.
Cell lines and transfection
The cervical epithelial tumour SiHa cell line (ATCC HTB-35), harbouring the HPV16 genome, and the SV40-transformed African green monkey kidney Cos-7 cells (ATCC CRL 1651), were used in this study. Both the SiHa and Cos-7 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat-inactivated fetal calf serum, 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. Transient transfection of 80% confluent cells was performed using the recombinant scFvExpress plasmids and Lipofectamine 2000 (Invitrogen) according to the manufacturer's recommendations. ScFv expression in SiHa cells was analysed 48 hours post-transfection by Western blotting as described below.
Analysis of the scFv expression
The identity of the scFvs expressed in prokaryotic systems was evaluated after separation on SDS-PAGE and blotting onto PVDF membrane, by Western blot analysis using the anti-FLAG M2 monoclonal antibody (mAb; Sigma, St. Louis, MO) as a primary antibody followed by a goat anti-mouse horseradish peroxidase-conjugated (GAM-HRP)) IgG (Amresco, Solon, OH).
Analysis of scFv expression in eukaryotic systems was performed by Western blotting after separation by SDS-PAGE of cell lysates in 2 × SDS loading buffer and blotting onto PVDF membrane. The rabbit anti-c-myc mAb, clone 9E10 (Sigma) was used as a primary antibody followed by a GAR-HRP-conjugated IgG (Cappel, Amresco, Solon, OH) incubation. The immunocomplexes were revealed by Chemiluminescence using the Super Signal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL).
Preparation and analysis of the soluble and insoluble extracts
Soluble and insoluble cellular extracts were obtained as described in ref. [
11]. Briefly, a 100 mm tissue culture dish of Cos-7 cells, transfected with each of the scFvE-cyto recombinant vectors, was lysed with 300 μl ice-cold extraction buffer containing containing Tris-Cl 20 mM, pH 8, MgCl2 20 mM, 0.5% NP40, 0.1 mg/ml leupeptin, chymostatin, and aprotinin, and 0.1 mM PMSF for 15 minutes. Cellular extracts were centrifuged at 15,000 rpm for 15 minutes at 4°C to separate soluble (supernatant) from insoluble proteins (pellet). Twenty μl of protein A-Sepharose in Tris-Cl 25 mM, pH 8.6, NaCl 150 mM (TBS) were incubated ON at 4°C with 2 μg of rabbit anti-c-myc mAb clone 9E10, washed three times with TBS, pH 8.0, and centrifuged for 1 minute at 1500 rpm. Three hundred μl of the soluble pool were incubated with 20 μl of 9E10-protein A-Sepharose for 2 hours at 4°C. The Sepharose beads were washed four times with TBS containing 0.1% NP40 and once with Tris-Cl 5 mM, pH 7.6. After addition of sample buffer (Tris-Cl 125 mM, pH 6.8, SDS 1%, glycerol 5%, DTT 10 mM, and bromophenol blue 0.005%), boiling for 2 minutes, and centrifugation for 1 minute (15,000 rpm, RT), the supernatants were loaded onto a 10% SDS-polyacrylamide gel and analysed by Western blotting as above described.
Discussion
In this paper we have analysed and compared the biophysical properties of three anti-16 E7 scFvs (scFv 32, 43 and 51) in view of their possible use
in vivo for therapy of the HPV-associated lesions. These scFvs exhibited different yield and solubility when expressed in prokaryotic system. ScFv43 resulted to be the least soluble, with the major part of expressed product remaining inside the bacterial cell cytoplasm, and just a small amount secreted into the periplasm. To achieve a sufficient production, different strategies were adopted. The scFv43 was extracted from the total cell lysate whereas the other two antibodies were extracted from the periplasm. The purified scFvs produced by the different procedures exhibited a good reactivity against the recombinant antigen [
18].
The thermal stability is a very important characteristic of the scFvs in view of their possible use in vivo. ScFv 43 and 32 have shown the same behaviour during the first hour of incubation at 37°C but the scFv43 was completely loosing its reactivity during the second hour of incubation, while the scFv32 reactivity, after the initial drop, remained stable up to 6 h. The scFv51 was the most stable antibody fragment because maintained its initial reactivity at least up to 72 hours of incubation.
In spite of its low thermal stability, the scFv43 seemed to work efficiently in hampering the E7 activity in the HPV16-positive SiHa cells, as we have reported in a previous study [
18]. These apparently conflicting results prompted us to investigate the reasons of this instability. The scFv stability depends on sequence contributions of both the framework and the CDRs, which are responsible for the antigen binding. Also, a mutation in one single residue of the framework can play an important role in maintaining the structure of the whole molecule [
30]. Many strategies have been adopted to improve stability of intrabodies for extracellular and intracellular applications [
31‐
33].
The comparison of the deduced scFv amino acid sequences with both the amino acid sequences of the Igs from the IMGT Database and the stable intracellular scFv consensus sequences of the VIDA database [
20] revealed some mismatching. At position VH5 of all the three scFv sequences, a sense mutation is present (Leu to Val, both neutral amino acids, numbering according to ref. [
25]).
At position VH34, scFv32 and scFv43 have a sense mutation which causes a change from a neutral to a polar amino acid (Met to Thr). ScFv43 presents additional mismatching at residues VH73 (Asp to Asn, both hydrophilic amino acids) and VL92 (Gly to His, from a neutral to a basic amino acid).
ScFv51 is the antibody fragment that shows the highest thermal stability and this result is in agreement with the identity of its framework sequences to the IgG sequences from the IMGT Database.
It is interesting to note that the VH34 residue is positioned in the Ig common core [
26]. The substitution of a hydrophobic with a hydrophilic amino acid in this region, as is the case of scFv32 and scFv43, may cause modifications in the molecule conformation. However, when attempting a correlation between the differences observed in the thermal stability of these two scFvs and their amino acid sequences, one has to consider that the VL regions of the two scFvs are derived from different germ-line genes.
Of interest, the scFv43 amino acids at positions VH 34 and 73 are not "fitting" the consensus sequence of the VIDA database, based on the scaffold of intrabodies with improved solubility and expression properties
in vivo [
20,
21]. A mutagenesis strategy has been designed to possibly improve the scFv43 stability and solubility by reverting the amino acids back to those present in the consensus sequence of both the IgG and the VIDA databases. The final aim was to improve the scFv43 intracellular performance without affecting its antigen-binding activity.
The amino acid at position VL92 resulted not to be important for the scFv43 stability. In contrast, reverting position VH34 seems to have fulfilled the goal, with scFv43 M2 showing improved features of thermal stability with respect to scFv43, as assessed by the shifting of t1/2 from 30 minutes to 24 hours.
Conclusion
In summary, the specific anti-16E7 scFvs described in this paper exhibit characteristics which are important for their functionality in eukaryotic systems. The thermal stability of scFv43, an antibody fragment able to contrast the E7 activity, has been improved, and we could expect scF43 M2 to function better than the parental scFv43. The scFv51 shows a high stability, a property often more important than affinity for the antigen in determining intracellular behaviour and efficacy, together with a high solubility. These properties render the two scFvs best candidates to be tested for anti-E7 activity in vivo.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
GMD participated in the study design, acquisition and interpretation of data and revision of the manuscript. CG participated in the interpretation of data and revision of manuscript. LA participated in the study design, acquisition and interpretation of data, coordinated the study, and drafted up the manuscript.
All authors read and approved the final manuscript.